RUTA:MED – Dual Workflow Medical Speech Transcription Pipeline and Editor
- Authors
- Znotiņš, Artūrs ; Darģis, Roberts ; Grūzītis, Normunds ; Bārzdiņš, Guntis and Goško, Didzis
- Item identifier
- http://hdl.handle.net/20.500.12574/70
- Demo URL
- https://med.ailab.lv/demo/
- Referenced by
- https://zenodo.org/record/6988837#.Y4JFaOzP1zU
- Date issued
- 2022
- Type
- toolService
- Language(s)
- Latvian
- Description
- In the medical domain, various approaches are used to produce examination reports and other medical records. Depending on the language-specific technology support, the type of examination, the size of the hospital or clinic, and other aspects, the reporting workflow can range from completely manual to (semi-)automated. A manual workflow may completely depend on the doctor itself or may include a transcriptionist centre in the loop. In an automated workflow, the transcriptionist centre is typically replaced by an automatic speech recognition (ASR) system. While the latter approach is well suited for high resource languages with low word error rate (WER), a dual approach combining automated transcription with the support from a transcriptionist centre may be more suited for less resourced languages. The RUTA:MED platform supports both workflows simultaneously. It currently includes an ASR pipeline for the Latvian language, which can be changed with ASR modules for any other language.
- Publisher
- AiLab IMCS UL
- Acknowledgement
-
European Regional Development Fund
Project code: 1.1.1.1/18/A/153
Project name: Latvian Speech Recognition and Synthesis for Medical Applications
- Collection(s)
-
